Hello list

A month or so ago I noticed that my server was freezing up. It would
slowly recover, but there were 10 or so entries like this:

kernel: swap_pager: indefinite wait buffer: bufobj: 0, blk
no: 140947, size: 32768
Feb 10 13:26:56 shell kernel: swap_pager: indefinite wait buffer: bufobj: 0, blk
no: 101732, size: 4096
Feb 10 13:32:57 shell kernel: swap_pager: indefinite wait buffer: bufobj: 0, blk
no: 86705, size: 32768
Feb 10 14:00:19 shell kernel: swap_pager: indefinite wait buffer: bufobj: 0, blk
no: 94369, size: 16384

Eventually, /var would fill up as well as swap and the system would
lock solid. I thought maybe there was a bad sector on swap, so played
around a bit with deleting and re-creating swap files, even made an
aux swap file and deleted the two real ones, to no avail. I upgraded
from freebsd-4.11 to 5.4 and then up to 6.1-prerelease and the problem
was still there. gstat(8) was still showing queued requests in the
leftmost column. The disk is an ibm deskstar. So I downloaded the
utilities software and ran strenuous testing against the disk, which
reported no errors.

Everywhere I looked on the lists, it pointed to bad cable or hardware.
The disk showed in dmesg like this:

Feb  9 21:22:41 shell kernel: ad0: 39266MB <IC35L040AVER07 0 ER4OA44A> at ata0-m
aster UDMA100

I thought I'd try running it in safe mode. dmesg now shows the disk like this:

ad0: 39266MB <IC35L040AVER07 0 ER4OA44A> at ata0-master PIO4

and no errors!

Is this a bug with the ata driver or a problem with the device?

--
John
_______________________________________________
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"

Reply via email to